B O T 
fubjedt to very great variation. The number of the parts 
is more liable to variation than their figure, and is found 
fometimes to vary even upon the fame plant; as in rue, 
golden f’axifrage, bird’s-neft, fpindle-tree, fyringa, and 
mofchatel; in the flowers of all which the number of the 
parts varies from five to four. In thefe doubtful cafes, 
the natural number mud be collected from the primary 
flower ; but in the variations of the number of the parts, 
there is a proportional affinity worth remarking : in flow¬ 
ers the (lamina ufually vary • from ten to eight, and from 
five^to four ; the corolla and calyx from five to four; and 
the whole flower from four to three : and the fruit alio 
ufually varies from five to three, and front five to four. 
The fituation of the parts is the mod condant, very rarely 
varying in the fame genus. The regularity of the petals 
is not fo much to be depended on as fonte former botanids 
have thought; for we fee in geranium, the European fpe- 
cies have regular corollre, but the African ones irregular. 
The nectarium nature has made of great confequence in 
.determining the genus. It is a decilive mark in all the 
following genera, viz. in lizard-flower, butterfly-flower, 
bird’s-ned, fumitory, violet, Malpighia, Banideria, bafe 
flower-fence, day-flower, bay-tree, white dittany, bean 
caper, marfh gentian, lily, crown imperial, water-leaf, 
crowfoot, Hermannia, barberry, bladder-nut, paffion-flow- 
er, daffodil, lea daffodil, marvel of Peru, rofe-bay, African 
fwallow-wort, African fpiraea, bell-flower, lead-wort, hya¬ 
cinth, dwarf rofe-bay, muftard, Kiggelaria, balfam-tree,co¬ 
lumbine, fennel-flower, wolf’s-bane or monk’s-hood, grafs 
of Parnaffus, barren-wort, chocolate-nut-tree, mignonette, 
Grevvia, hellebore, ifopyrum, Indian crefs, and balfam. 
The flamina and calyx, being lefs fubjedt to luxuriancy, 
are far more certain than the petals in deciding genera. 
The corolla varies as to its figure in many genera; as in 
whortle-berry, winter green, bafe heath, tobacco, bog-bean, 
primrofe, fpeedwell, gentian, hyacinth, fcabious, and daffo¬ 
dil. It varies alfo as to number, being in ranunculus pen¬ 
tapetalous in fome fpecies, and polypetalous in others; in 
hellebore alfo, pentapetalous and polypetalous ; in thrift, 
pentapetalous and monopetalous; and in fumitory, diape- 
talous and tetrapetalous: and the number is alfo fome¬ 
times variable in the fame fpecies; as is obferved in the 
papaw-tree, and caffada. The ftrufture of the pericarpium 
•was formerly thought to be of great confequence in deter¬ 
mining the genera ; but there are examples without num¬ 
ber that demonftrate its uncertainty. 
The characters of luxuriant flowers, whether eunuchs, 
i. e. fuch as have loft the (lamina, or mutilate, cannot be 
allowed any place in determining the genera ; for in full 
flowers no number of petals can be affigned, and the (la¬ 
mina are generally wanting, the number of which makes 
a principal part of the generic character ; and in mutilate 
flowers, as in fome fpecies of bell-flower, quimoefit, and 
ruellia, the corolla would be excluded from the deferip- 
tion, contrary to the nature of the other fpecies of the ge¬ 
nus. But as the calyx in full flowers is fcarcely ever al¬ 
tered, it may detedt the genus; and the lowed feries of 
petals in polypetalous corollse remaining the fame in refpedt 
to number, the genus may alfo be often known by that 
Character ; as in the rofe, the poppy, and fennel-flower. 
Of SPECIES. 
Species conftitute the third fubdivillon in the Linnaean 
fyftem, which comprehends all the different forms of plants 
that are fuppofed to have been originally created. Thefe 
plants, fays Linnaeus, have, by the eftablilhed laws of na¬ 
ture, continued to produce others like themfelves; there¬ 
fore the fpecies plantarum comprehends all the different in¬ 
variable forms of plants which are found at this day upon 
the face of the earth. 
We have (hewn the generic diftinCtions to depend on the 
form of the fructification, and to be confined to that alone. 
Specific differences take their rife from any circumftance, 
wherein plants of the fame genus are found to difagree ; 
provided fuch circumftance is conftant, and not liable to 
Vol.III. No. 129. 
ANY. _ 2 6t 
alteration by culture or other accidents ; for all cafual dif¬ 
ferences are varieties of the fame fpecies. Towards the end 
of -the leventeenth century, the defire of increaling the 
number of plants had fo (eized the botanifts of that time, 
that new fpecies were eftablifhed on too (light differences, 
to the great detriment of the fcience ; and rite fame eager- 
nefs led them alfo to fet down as new genera, .what (liould 
have been fpecies only. This evil was in fome mealure 
unavoidable, vvhilft there were no fixed principles for the 
regulation-of the fcience in this refpeCt. A remedy for it 
was firtl attempted by Vaillant; afterwards by Juffieu, 
Haller, Royenus, Gronovius, and others; and laflly by 
Linnaeus, vvhofe aphorifms have brought the work much 
nearer to perfection. Something indeed feems (till want¬ 
ing to complete thefe doCtrines; but perhaps more is not 
to be expected till this branch of natural philofophy re¬ 
ceives farther affiftance from experiment. We (hall firff 
point out thofe circumftances by which fpecies are diftjn- 
guidied with certainty, and then (hew what are varieties. 
The root will fometimes afford areal fpecific difference, 
as in fquill, where the fpecies are fcarcely to be diftinguifti- 
ed, but by the bulbs being tunicate, folid, or fquamofe ; 
and in the butterfly-flower, where the fpecies are known 
by the roots being fibrous, round, or tefficulate ; but as 
accefs cannot always be conveniently had to this part of 
the plant, it is better to fix the fpecific diftinction on fome 
more obvious circumftance. The trunk often furnifties a 
fure mark ol diftinction. Thus in' St. John’s wort, lily of 
the valley, and French honeyfuckle, there are many fpe¬ 
cies diltinguiffiable by the angles of the (lent; and in lupine, 
the (pecies are not eafy to be known, except by the fame 
part being Gmple, or.compound. In net-wort, the mod 
remarkable difference is in the (tern, which is quinquangu- 
lar, hexangular, decangular, See. In winter-green, fome 
fpecies are diftiiiguiftied by a triquetrous (talk. In citrus, 
the aurantium is diftinguiftied from its congeners by its 
petioles, which are winged or inerdafed by a membrane on 
each fide; and in globe amaranth, there is a fpecies diftin¬ 
guifhed by its peduncles, which are diphyllous, being fur- 
nidied with two oppofite folioles which are placed under 
the head of the flowers. 
The /eaves exhibit 1110ft natural and alfo mod elegant fpe¬ 
cific diff erences. Thefe have been fo amply treated of al¬ 
ready, that it would be only repetition to particularize or ex¬ 
emplify the numerous cafes that occur of fuch diftindtions. 
The fulcra, props or defences of plants, are generally good 
marks of diftindtion, and muff be carefully attended to by 
the botanift for the determination of the fpecies; efpecially 
where the difference confifts principally in thofe parts of 
the plant. Thus, the aculei, or prickles, are remarkable in 
the rafpberry ; the fpines, or thorns, in the plum-tree and 
hawthorn; the bradtcce, or floral leaves, in fumitory, dra- 
gon’s-head, and the Indian fpecies of French honeyfuckle ; 
to which may be added the coma, which is a buffiy head, 
compofed of bradtete that are of a large fize, and terminate 
the (tern in crown imperial, lavender, and fuge. Glandules , 
or the fecretory glands, furnifti the elfential mark in Indian 
mallow, fenlitive plant, wild fenna, and many other genera. 
Thefe glands are found on tire ferratures at the bafe of 
the leaves, in fun-feed, willow, and the almond-tree ; on 
the back of the leaves, in Indian mallow, and tire pallion- 
flower; and on the aculei, in Bauhinea aculeata, where by 
the apex of the aculei a liquor is fecreted. The almond- 
tree is diftinguiftied from perfica only by the glandules of 
the ferratures; nor could t lie fpecies of the Indian mallow 
be ever fixed without examining the glandules of the leaves. 
The convolvulus with a tuberculate calyx, is fo variable 
in the (liape of its leaves, that it feems diviiible into many 
fpecies, yet it is kept together by the glandules: and there 
is a fpecies of Ofwego tea, diftingui(liable from its conge¬ 
ners by the glandules, that are fprinkled over the corolla. 
The fipulcc, or fmall fcqly leaves at the bafe of a leaf- 
(talk or fruit-ftalk, are of great confequence in many ex- 
tenfive genara, where the fpecies are liable to confulion. 
Thus in one fpecies of the honey-flower the ftipuhe are 
3 X folitary 5 
